Consider a disease that accounts for hundreds of thousands of hospital admissions and tens of thousands of deaths per year. A disease that, despite our best efforts, still boasts a mortality rate of at least one in every ten patients afflicted. Community-acquired pneumonia (CAP) is just that disease and it is the leading cause of death by infection.
In this issue, Dr. Alfred Gin provides us with a succinct review of the issues and controversies surrounding the management of CAP. As Gin describes, CAP has been the subject of considerable attention over the past decade. Despite this interest, many issues persist. Streptococcus pneumoniae remains the most common pathogen causing CAP; however, treatment has become more complex as the susceptibility patterns for this organism are changing. The incidence of S. pneumoniae with some degree of resistance to penicillin has increased three-fold in the US over a 7-year period and about five-fold in Canada over the last decade. While true in vitro resistance appears to be relatively uncommon (and possibly declining) in Canada, conditions in the U.S. are somewhat worse. Why our resistance patterns are different to those of our southern neighbours is still unknown. Perhaps it is related to the relative restrictions and use of antimicrobials. No doubt a number of factors are at work here. Controversy also exists regarding the most appropriate breakpoints to describe sensitivities for S. pneumoniae. Recommendations have apparently been made to increase the MIC breakpoints that we use to define resistance. Doing so will reduce the reported prevalence of penicillin resistance and we will again need to reconsider our treatment choices. Gin also reminds us that S. pneumoniae resistance to fluoroquinolones appears to be linked to first-generation agent utilization rates. A relationship we should keep in mind when considering the new guidelines.
Annual CAP-related costs in the US are estimated to be $8 billion. If we apply the usual 10% rule to account for population size differences between the U.S. and Canada, that would suggest that the annual costs in our country are approximately $800 million. Needless to say, the costs to society (and the potential profits for the pharmaceutical industry) associated with the treatment of CAP are substantial. It is no wonder that this disease has captured the attention of the pharmaceutical manufacturers and that there are so many agents available in the marketplace to treat this affliction.
Most importantly, Gin alerts us to the recent publication of the Canadian Guidelines for the Initial Management of CAP. After a 7-year hiatus, these guidelines have been updated and, for the first time, are now available on line. While not explicitly stated, these guidelines are directed at the initial management of CAP in adult patients only. Little emphasis is placed upon the continued treatment and monitoring of patients who either respond to or fail on initial therapy. In addition, no information regarding pneumonia in pediatric patients is provided.
The guidelines were created by a group who denote themselves the Canadian Community-Acquired Pneumonia Working Group, although some of the members are actually from the U.S. and abroad. To accomplish this task, select members of the Canadian Infectious Diseases Society (CIDS), the Canadian Thoracic Society (CTS) and additional members from the U.S. apparently separated into 3 teams who were individually responsible for addressing epidemiology, risk factors and etiology, and treatment. According to the authors, the teams met for two days in early 1999, formulated drafts, circulated these to the members, sought an external review by colleagues in the US, Europe and Israel and then finalized the guidelines based upon the input received. With the possible exception of one individual with dual Pharm.D./Ph.D. degrees, there was no apparent involvement or collaboration with individuals or groups outside of the medical discipline. Now that’s consensus building. Perhaps the members support the position of the Infectious Diseases Society of America (as written by the selective membership of the Clinical Affairs Committee and the Antimicrobial Use and Clinical Trials Committees) who promote a limited role of pharmacists in the management of infections? I truly hope this is not the case, as no one discipline would be fully functional without the collaboration of the balance. The guidelines would have been stronger if their creation had been a true interdisciplinary effort as well as international exercise. Certainly, we owe it to our patients to work synergistically on their behalf.
That said, I must applaud the authors of the guidelines for what must have constituted a significant amount of coordination and work. When properly done, practice guidelines can be a useful tool to help us translate research evidence into clinical decision aids. The creation of this particular comprehensive, well-referenced guidelines document aimed at influencing the management of CAP in Canada required an extensive and detailed systematic review of the literature. The authors state that they were required to review 585 published papers or abstracts for the treatment section alone. Whether this was accomplished with or without external sponsorship is unknown. Unfortunately, no information was provided by the journal or the authors regarding the source (or absence) of external funding for this worthwhile exercise.
Let’s add some further context to these guidelines. Although they have been denoted “Canadian guidelines”, it would be more accurate to simply describe these as “2000 CIDS/CTS Guidelines” for the management of community-acquired pneumonia, much as the Infectious Diseases Society of America has done with the publication of their recent practice guidelines. It would be an oversimplification to assume that these guidelines can be applied uniformly across this vast country with its inherent regional and local differences in bacterial susceptibility patterns, patient characteristics, access to health care and other factors. While the authors describe these new guidelines as “an evidence-based update”, they responsibly also state that “it is important to recognize that these recommendations are derived by the consensus of experts and not entirely based on evidence from randomized clinical trials”. I will try to illustrate this critical qualifying statement with an example.
One of the more significant changes to the list of empirical choices for the initial management of CAP is the addition of the so-called “respiratory fluoroquinolones”. These agents are apparently denoted as such because they possess enhanced in vitro activity against the pathogens typically responsible for CAP, in particular S. pneumoniae. I personally believe that the term “respiratory fluoroquinolone” was conceived as a marketing ploy aimed at associating this group of drugs with a particular type of infection. What a useful “handle” to promote a particular drug in the increasingly confusing milieu of fluoroquinolones. What other antimicrobial do you know that has been linked to a particular anatomical structure and its related infections? Do we denote cefuroxime a “respiratory cephalosporin” or erythromycin as a “respiratory macrolide”? Of the list of respiratory fluoroquinolones identified by the authors of the guidelines, grepafloxacin has been withdrawn from the market due to cardiac toxicity and trovafloxacin is restricted to use in the hospital setting due to reports of severe liver toxicity. Not a sterling track record for the new fluoroquinolones. Neither gatifloxacin nor moxifloxacin are as yet available in Canada. Although available, ofloxacin is mysteriously missing from the text (perhaps the authors incorrectly took to heart the suggested synonyms supplied by MS Word when spell checking this word). That’s a shame. Unfortunately, that leaves us with only one agent, levofloxacin.
According to the new guidelines, levofloxacin has now been elevated to the role of drug of first choice for the treatment of CAP in patients with chronic obstructive lung disease who have received antibiotics or corticosteroids within the past 3 months; nursing home residents or hospitalized patients in the non-intensive care unit setting; and as second line therapy in patients with suspected macroaspiration (in combination with clindamycin or metronidazole). Let’s briefly examine the evidence behind this selection.
While the authors describe that there are at least 16 published clinical trials involving comparisons of monotherapy with various respiratory fluoroquinolones against other fluoroquinolones, macrolides and cephalosporins, seven of these are in abstract form only. Until these undergo appropriate peer-review, these publications should be considered research findings only and should be used with great caution when considering the management of a particular patient, let alone the creation of national guidelines. According to the guidelines authors, only three clinical trials have been published in which levofloxacin has been directly compared to standard therapy for CAP. Two of these studies have undergone peer-review and have been published in reputable journals. The third report is a conference abstract for which there appears to be no follow-up full report. Again, while the study is large, the information has not been subjected to appropriate peer-review.
The two published peer-reviewed studies involved industry-sponsored open-label or unblinded study designs in which levofloxacin was compared to a variety of cephalosporin regimens. The study by File et al. involved 590 patients of which 456 (226 parenteral and/or oral levofloxacin 500mg once daily, 230 ceftriaxone 1g or 2g Q12-24H and/or cefuroxime axetil 500mg Q12H) were considered by the authors to be evaluable for clinical efficacy. The decision to use an intravenous or oral antimicrobial agent for initial therapy was made by the investigator. Erythromycin or doxycycline could be added to the comparator arm at the investigator’s discretion. Remember, this is an unblinded trial. No intention-to-treat analysis was apparently conducted. Clinical success at 5-7 days post therapy was reported to be superior for the levofloxacin group (96%) as compared to the ceftriaxone and/or cefuroxime axetil group (90%) (95% confidence interval [CI] of -10.7 to -1.3). Whether anyone would consider a 1% difference in clinical outcomes to be superior is debatable, particularly in light of the variability in regimens employed and the unblinded nature of this company-sponsored trial. In this particular study, there was a relatively low incidence of S. pneumoniae and Haemophilus influenzae (isolated in 15 and 12% of clinically evaluable patients, respectively) and a relatively high incidence of atypical organisms. These are conditions that would favour antimicrobials (or combinations if employed) with activity against the latter pathogens. The second unblinded study by Norrby et al. involved 625 patients who were enrolled and randomized to receive parenteral and/or oral levofloxacin (500mg Q12H) or ceftriaxone (4 g/d). Of these, 6 patients received no treatment and the remaining 619 (levofloxacin 314, ceftriaxone 305) patients constituted what the authors describe as their intention-to-treat group. At the end of treatment there were only 266 patients (127 levofloxacin, 139 ceftriaxone) who had completed the study protocol. At the clinical endpoint (2-5 d after the end of treatment), the cure rates for levofloxacin and ceftriaxone were similar in both the intention to treat analysis (76% and 75%, respectively) and completed protocol analysis (87% and 86%, respectively).
To summarize, we have two industry-sponsored, unblinded and methodologically dissimilar clinical trials involving about 1,200 enrolled patients in which variable regimens of parenteral or oral levofloxacin were compared to variable regimens of two cephalosporins (at the discretion of the investigators). Levofloxacin appears to be superior to its comparators in one trial and equivalent to the comparator in the other. Consider this evidence when you apply the guidelines to the many thousands of patients we will treat for CAP every year. Once again, the qualifying statement by the authors of the guidelines ring true. This particular recommendation (and some others in the document) was based upon consensus opinion and was made in the absence of an abundance of solid evidence. Combine this with the fact that, like any other drug class, the increased use of respiratory fluoroquinolones can be expected to lead to an increased incidence of bacterial resistance and we have an interesting conundrum. As pointed out by Gin, preliminary reports of respiratory fluoroquinolone resistance-related treatment failures have already surfaced this year. Time will only tell how long the clinical utility half-life of the newer fluoroquinolones will be. As aptly recommended by Hooper, in order to preserve the therapeutic opportunities that the most recent generation of fluoroquinolones provide, our challenge will be to ensure that we use these drugs wisely. Most community-acquired respiratory infections are still caused by pathogens that remain susceptible to beta-lactam drugs and macrolides. Fluoroquinolones should only be used in clinical situations where they offer a clear therapeutic advantage over other drug classes.
There are many other issues to consider when treating CAP, but I’ll close with one only. As described by Cook et al, not all the outcomes of interest are measured in all primary studies. Certain measures, such as quality of life, are often inadequately represented in reviews. This can be no better illustrated than in the context of CAP management. While we should continue to strive to eradicate the organism, clear the x-ray and reduce the fever…what definitive evidence do we have that we have actually improved the patient’s quality of life, let alone reduced mortality?
In summary, while the newly published guidelines certainly provide us with a very useful summary of our present understanding of the epidemiology, risk factors, etiology, diagnosis and treatment of CAP, they do not represent new knowledge. What is the take home message? Apply these guidelines where appropriate, but not without an understanding of their limitations. As Cook also describes, “…no matter how they are developed and implemented, guidelines may do more harm than good if they are inappropriately interpreted or applied”.